Well guys ...
It's like this. You ask for information in hopes an expert in the field answers with the technically correct answer using the industry accepted materials and supplies. Don't turn around and complain that you don't like the answer. That's what many of you are doing.
Myself and others have also given alternatives and guidelines, and you can't complain about being that open about things. The sad truth is that the amateur tech is the one who will believe the meter is accurate to all visible digits (admit it) when some of those meters a lying about the last two digits. Then, measurements on low resistance items like speaker voice coils or load resistors, even at 8 ohms, with two leads. No consideration given for temperature effects of the meter and it's internal voltage reference or thermocouple effects from dissimilar metallic contacts. Did anyone who made measurements of heating elements state the ambient temperature and load temperatures? Where the measurements made two wire, or Kelvin connected? Kelvin leads or at least two lead sets? How about the meter make and model listed?
None that I have seen. A professional would create an error budget for everything and combine the results to give a figure to a realistic number of significant digits, then add the uncertainty term as + or - x ohms. I'll state the obvious for some of us. These resistance measurements need to be made with a 4 wire (Kelvin) connection to begin with. If not, your accuracy is so low that you may as well just quote the spec as worst case (the most error you can find). Then figure out the tempco and attempt to measure output power.
Nasty surprise time. How far up in frequency is your true rms meter good for and at what spec? Most meters are in trouble from 200Hz up. Mine is good to 100 KHz, and the bench meter for 300 KHz - exceptional performance. The LCR meter I use will go up to 100 KHz test frequency as well. It also uses Kelvin clips (4 wire).
Many of you will still quote the power to one decimal at least using a meter outside its bandwidth, peak indicating. A hopelessly indefensible reading. But, by gosh, that is the number they will stick to.
If you want to get a rough idea of output, or you simply want to heat an amplifier up to set or check bias current stability - fine. If you are preparing to quote a number, you had better have done your homework and used good testing and measurement practices to arrive at a number that can be verified by someone else should the need ever arise.
Like it or not, that is the answer. Proper resistors are not that expensive and will last a lifetime (barring accidents). Either get those, or quote your numbers as very rough, or imprecise. At the bottom of all this is a meter that is half decent. Many are available used in various places, even new decent ones starting around $220~250 (Keysight or Fluke). but read the specs for the different models so you know what you are giving up with a lower cost meter. That meter will also last you a lifetime - mine have so far. (new in 1968).
-Chris
It's like this. You ask for information in hopes an expert in the field answers with the technically correct answer using the industry accepted materials and supplies. Don't turn around and complain that you don't like the answer. That's what many of you are doing.
Myself and others have also given alternatives and guidelines, and you can't complain about being that open about things. The sad truth is that the amateur tech is the one who will believe the meter is accurate to all visible digits (admit it) when some of those meters a lying about the last two digits. Then, measurements on low resistance items like speaker voice coils or load resistors, even at 8 ohms, with two leads. No consideration given for temperature effects of the meter and it's internal voltage reference or thermocouple effects from dissimilar metallic contacts. Did anyone who made measurements of heating elements state the ambient temperature and load temperatures? Where the measurements made two wire, or Kelvin connected? Kelvin leads or at least two lead sets? How about the meter make and model listed?
None that I have seen. A professional would create an error budget for everything and combine the results to give a figure to a realistic number of significant digits, then add the uncertainty term as + or - x ohms. I'll state the obvious for some of us. These resistance measurements need to be made with a 4 wire (Kelvin) connection to begin with. If not, your accuracy is so low that you may as well just quote the spec as worst case (the most error you can find). Then figure out the tempco and attempt to measure output power.
Nasty surprise time. How far up in frequency is your true rms meter good for and at what spec? Most meters are in trouble from 200Hz up. Mine is good to 100 KHz, and the bench meter for 300 KHz - exceptional performance. The LCR meter I use will go up to 100 KHz test frequency as well. It also uses Kelvin clips (4 wire).
Many of you will still quote the power to one decimal at least using a meter outside its bandwidth, peak indicating. A hopelessly indefensible reading. But, by gosh, that is the number they will stick to.
If you want to get a rough idea of output, or you simply want to heat an amplifier up to set or check bias current stability - fine. If you are preparing to quote a number, you had better have done your homework and used good testing and measurement practices to arrive at a number that can be verified by someone else should the need ever arise.
Like it or not, that is the answer. Proper resistors are not that expensive and will last a lifetime (barring accidents). Either get those, or quote your numbers as very rough, or imprecise. At the bottom of all this is a meter that is half decent. Many are available used in various places, even new decent ones starting around $220~250 (Keysight or Fluke). but read the specs for the different models so you know what you are giving up with a lower cost meter. That meter will also last you a lifetime - mine have so far. (new in 1968).
-Chris
Hi Dimitri,
Okay, your 2100 doesn't measure resistance it seems, other specs are:
Valhalla 2100 Digital Power Analyzer
dc & 40 Hz to 5 kHz: ±0.1% of reading ±6 digits
5 kHz to 15 kHz: ±0.5% of reading ±6 digits
15 kHz to 20 kHz: ±0.75% of reading ±6 digits useable to 50 kHz
The 2101:
Valhalla 2101 Digital Power Analyzer
dc & 40 Hz to 5 kHz: ±0.1% of reading ±6 digits
5 kHz to 15 kHz: ±0.5% of reading ±6 digits
15 kHz to 20 kHz: ±0.75% of reading ±6 digits useable to 50 kHz
Again, no resistance reading, but both meters have a basic accuracy of 1% on power measurements, plus similar errors like the voltage readings. Your lowest range is 30 V.
They are true RMS reading meters but are really designed for AC power mains use. I'd have to say that you probably get some better numbers than many members, but run your numbers to generate your error amounts.
-Chris
Okay, your 2100 doesn't measure resistance it seems, other specs are:
Valhalla 2100 Digital Power Analyzer
dc & 40 Hz to 5 kHz: ±0.1% of reading ±6 digits
5 kHz to 15 kHz: ±0.5% of reading ±6 digits
15 kHz to 20 kHz: ±0.75% of reading ±6 digits useable to 50 kHz
The 2101:
Valhalla 2101 Digital Power Analyzer
dc & 40 Hz to 5 kHz: ±0.1% of reading ±6 digits
5 kHz to 15 kHz: ±0.5% of reading ±6 digits
15 kHz to 20 kHz: ±0.75% of reading ±6 digits useable to 50 kHz
Again, no resistance reading, but both meters have a basic accuracy of 1% on power measurements, plus similar errors like the voltage readings. Your lowest range is 30 V.
They are true RMS reading meters but are really designed for AC power mains use. I'd have to say that you probably get some better numbers than many members, but run your numbers to generate your error amounts.
-Chris
Oh wow.
I think I am beginning to understand.
The operative word...beginning.
Gotta list the 'quipment and the conditions.
Cheers,
Hi cbdb,
If you read my earlier posts, and others, you will see that alternatives were freely given. The OP was simply told what conditions would apply so he didn't go and make a fool of himself if he presented an approximate answer as an exact one.
The worst thing you can do to someone is give them a partial answer when they need all the information. He has stated he expects to do more work on amplifiers in the future which I think you missed.
I have a scary amount of money invested in equipment and parts. Replacement cost will exceed $100,000. Sometimes I do fix a $100 amplifier. Should I, or should I only work on expensive stuff? If you were building a $100 amplifier, would it be okay if your $25 meter lied to you, causing damage that you could never figure out? This situation has happened more often than you know. If you like this hobby, you can start out with less expensive equipment. But sooner or later you will have to buy better things to be able to do the work you want to do. Guys who collect stamps know this, as do butterfly collectors and fish collectors. Every hobby has a buy in as you progress.
-Chris
If you read my earlier posts, and others, you will see that alternatives were freely given. The OP was simply told what conditions would apply so he didn't go and make a fool of himself if he presented an approximate answer as an exact one.
The worst thing you can do to someone is give them a partial answer when they need all the information. He has stated he expects to do more work on amplifiers in the future which I think you missed.
I have a scary amount of money invested in equipment and parts. Replacement cost will exceed $100,000. Sometimes I do fix a $100 amplifier. Should I, or should I only work on expensive stuff? If you were building a $100 amplifier, would it be okay if your $25 meter lied to you, causing damage that you could never figure out? This situation has happened more often than you know. If you like this hobby, you can start out with less expensive equipment. But sooner or later you will have to buy better things to be able to do the work you want to do. Guys who collect stamps know this, as do butterfly collectors and fish collectors. Every hobby has a buy in as you progress.
-Chris
Thank you cbdb and once again thanks anatech. I'm sorry for such a late response on this thread.
I think I will do some specific tests with meters of mine and the two old heating elements I already have. I will use a shunt and free air, not in a bucket of water so I can get the biggest temperature changes and swings. If the shunts voltage dif changes then that means the resistance would change- correct? I will use general 60hz. That way its constant, I don't need to worry about changing to the meter or the accuracy of the resistance of my fluke 179.
For you satisfaction though, anatech, I am going to get a set, correction- I will order a set of these resistors tonight(!) capable of testing only one channel at a time 4 of these " 8 OHM 8R 200W Watt Power Metal Shell Case Wirewound Resistor | eBay "
The purpose for testing is both stress and an accurate power out. I want to finally engrave the numbers on my 1/4 inch plate aluminum for my meters. They meters bounce but I never know how much its playing at. The amps internal voltage is +/- 85 volts, the transformer is pretty much acting as an isolation transformer in the setup. I am designing a new mainboard with an extra gain step in there to help drive the output of it. I have 20 output transistors per side 10 pnp 10 npn complementary (40 in total!). I have an overrated soft start circuit and a fan blower. Current load for the amps is 2 ohm. Refer to an earlier post for a video....
Just a note, to replace all the outputs in the amplifier would cost a little less than $$$600 :O
Thank you to anyone else who responded to the tread again! I appreciate the input from all!
I think I will do some specific tests with meters of mine and the two old heating elements I already have. I will use a shunt and free air, not in a bucket of water so I can get the biggest temperature changes and swings. If the shunts voltage dif changes then that means the resistance would change- correct? I will use general 60hz. That way its constant, I don't need to worry about changing to the meter or the accuracy of the resistance of my fluke 179.
For you satisfaction though, anatech, I am going to get a set, correction- I will order a set of these resistors tonight(!) capable of testing only one channel at a time 4 of these " 8 OHM 8R 200W Watt Power Metal Shell Case Wirewound Resistor | eBay "
The purpose for testing is both stress and an accurate power out. I want to finally engrave the numbers on my 1/4 inch plate aluminum for my meters. They meters bounce but I never know how much its playing at. The amps internal voltage is +/- 85 volts, the transformer is pretty much acting as an isolation transformer in the setup. I am designing a new mainboard with an extra gain step in there to help drive the output of it. I have 20 output transistors per side 10 pnp 10 npn complementary (40 in total!). I have an overrated soft start circuit and a fan blower. Current load for the amps is 2 ohm. Refer to an earlier post for a video....
Just a note, to replace all the outputs in the amplifier would cost a little less than $$$600 :O
Thank you to anyone else who responded to the tread again! I appreciate the input from all!
Last edited:
Hi Gamerandds,
Cool, those resistors are really cheap!
The resistors are rated at 200 watts each. Mount one pair on a good sized heat sink and that should do you. If you use four of them per channel, you can dissipate 800 watts. That is a lot of power, 1,600 watts total power is breaker tripping territory. This should only cost you $30.40 USD plus a heat sink all in for 200 watts per channel. That is a very reasonable cost. The 250 watt Dale resistors I use cost $50 plus taxes and shipping each, plus the heat sinks and milling ($50 CDN each), and I made three sets. That was a substantial outlay for one guy (me). But, that was what I needed to satisfy the terms of most warranty contracts, or even to be able to quote any power figures.
-Chris
Cool, those resistors are really cheap!
The resistors are rated at 200 watts each. Mount one pair on a good sized heat sink and that should do you. If you use four of them per channel, you can dissipate 800 watts. That is a lot of power, 1,600 watts total power is breaker tripping territory. This should only cost you $30.40 USD plus a heat sink all in for 200 watts per channel. That is a very reasonable cost. The 250 watt Dale resistors I use cost $50 plus taxes and shipping each, plus the heat sinks and milling ($50 CDN each), and I made three sets. That was a substantial outlay for one guy (me). But, that was what I needed to satisfy the terms of most warranty contracts, or even to be able to quote any power figures.
-Chris
Exactly why I'm thinking of alternatives. The other thing is that I'm not a shop- so it's not like I get amps to fix and work on every day. Its just mostly for the odd amp here and there I fix.
I just bought these which seem to be a pretty affordable alternative. Don't know how stable they will be but I can test that and at this price not much to lose. Qnty 9 of these in series parallel will give close to 8R and almost 250w. I plan to mount to a fan cooled heatsink. Also, for smaller amps I can select 4R or 8R loads.
Online Shop 5Pcs 8ohm R 8 Ohm 8R 25W Watt High Quality Wirewound Aluminum Power Metal Shell Case Resistance Resistor Free shipping|Aliexpress Mobile
Related topic - what do you all use for a low cost way to do DIY THD measurements?
Online Shop 5Pcs 8ohm R 8 Ohm 8R 25W Watt High Quality Wirewound Aluminum Power Metal Shell Case Resistance Resistor Free shipping|Aliexpress Mobile
Related topic - what do you all use for a low cost way to do DIY THD measurements?
Last edited:
I know rod Elliot from sound.weathost.com did a similar thing your doing using regular ceramic wire wound resistor in a bucket of oil, all wired in a weird way... (Series parallel) he claims,and don't quote me on this, but at least 2 times the power and he has ran them up to a staggering 5x their power rating in oil- mark you for a short period of time-
The oil conducts heat away but is a mess. Actually deionized water is a great dielectric and good heat removal fluid but may have local boiling. Series parallel is just a convenient way to keep resistance same as single resistor but spread load. Can be done with square numbers of resistors like 4, 9, 16, 25. For cutting down to 4R then switch around for fewer in series and more in parallel.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.